2.4 The bolometric technique
2.4.4 Detector operation
In order to convert the resistance variation into a readable voltage signal, the sensor is polarized by the simple circuit shown in fig.2.7(a). A bias current IB is produced by a
voltage generator closed on a load resistance that is put in series with the thermistor. The load resistance RL is set much bigger than the sensor resistance R(T ), so that
the bias current can be assumed to be constant. In these conditions a voltage drop V (T ) = I · R(T ) appears across the thermistor, causing on it a power dissipation P = I· V . This in turn produces an increase of the sensor temperature and therefore its resistance is decreased. There is therefore an equilibrium condition when the power dissipated on the sensor equals the heat dissipation versus the thermal bath. Denoting by T0 the detector base temperature and by G the thermal conductance versus the
bath, the equilibrium temperature of the sensor is given by the expression Tbol = T0 +
P
G . (2.18)
The dependence of the sensor resistance on the dissipated power is depicted in fig. 2.7(b).
Because of the dependence of the resistance on the bias current, the characteristic I − V curve of the detectors deviates from linearity, giving rise to the non-ohmic behavior represented in fig.2.8(a). For small values of the bias current the temperature rise produced by power dissipation can be neglected and the I−V curve is almost linear. For bigger values of IBthe slope of the I−V curve starts to increase, until an inversion
point is reached, where a further increase of the bias current causes a decrease of the sensor voltage. As represented in fig. 2.8(a) the working point of the detector can be found by the intersection of the load curve with the load line imposed by the bias circuit: IB = (VB+ Vbol) /RL, where VB is the bias voltage. It is chosen to maximize
the signal amplitude or more precisely the signal to noise ratio. As represented in fig. 2.8(b) this usually correspond to a bias current at halfway between the end linear range and the inversion point.
+ VBIAS RL/2 Rbol(T) Ibol Vbol(T) RL/2 (a) 6 10 20 4 8 Temperature P 10-2 1 102 104 106 [pW] logR 2 1 (b)
Figure 2.7: The left picture shows the electric scheme of the bias circuit used for thermistor readout. The right picture shows the dependence of the resistance on the power dissipation for various values of the base temperature. Curves with lower resistance at P=0 correspond to higher base temperatures.
VBIAS Vbol Ibol Working Point (WP) Load Curve (LC) Load Line Inversion Point (IP) V BIAS R L (a)
I
0.8 0.4 1.2 2 2.4 2.8 3.2 b Optimum point[pA]
Signal [A.U.] 1.6 2 4 6 8V [mV]
10 Load curve (b)Figure 2.8: Load curves for semiconductor thermistors. On the left picture the working point is determined by intersection of the sensor characteristic curve with the bias circuit load line. On the right the load curve is shown together with the corresponding signal amplitude.
The signal amplitude produced by an energy release E can be estimated as follows. In static conditions the voltage over the sensor is given by
Vbol = VB
Rbol
RL+ Rbol
, (2.19)
where VB is the bias voltage of the polarization circuit. The voltage pulse produced
∆Vbol = VB RL (RL+ Rbol)2 ≃ C TE bol ApP Rbol , (2.20)
where eq. (2.7) and eq. (2.14) have been used. The signal amplitude in eq. (2.20) vanishes for both P → 0 and P → ∞, as for big power dissipation the logarithmic sensitivity and sensor resistance approach rapidly to small values.
A typical pulse produced by a particle interacting in a bolometers is represented in fig. 2.9. Using some numbers relative to the CUORE bolometers it is possible to have an idea of the magnitude of the produced signal. A typical value for the absorber heat capacity is C ≃ 10−9J/K at 10 mK, thus an energy release of 1 M eV would result in
a temperature rise of ∼ 0.1mK. Since the typical voltage drop across the sensor is of few mV in static conditions, the pulse height produced by the energy release of 1 MeV is given by ∆V /V ∼ ∆R/R ∼ A∆T/T ∼ 100µV .
400 800 1200 1600
Pulse amplitude [a.u.]
T [ms]
The
CUORICINO experiment
3.1
Introduction
The use of bolometric detectors for the search of neutrinoless double beta decay was proposed by Fiorini in 1984 [52]. Since more than twenty years his research group has been developing cryogenic detectors of increasing mass. The successful operation of a 340 g Tellurium dioxide crystal [57] was followed by the construction of a detector array composed by 20 crystals, for a total mass of 6.8 kg of TeO2 (MiDBD [58]).
A further mass increase was obtained with the recently completed CUORICINO ex- periment [32,59,60]. Operated in the Laboratori Nazionali del Gran Sasso in the years 2003-2008, CUORICINO was composed by a tower of 62 TeO2 bolometers, with a total
mass of ∼41 kg. The experiment was able to set a lower limit of 3.0 × 1024y for the
0νββ half life of130Te (m
ββ < 0.2÷ 0.68 eV ). At present the CUORICINO results rep-
resent one of the most competitive limits for the effective Majorana mass, comparable with the ones obtained with Germanium detectors [27, 38]. Furthermore, it has been a unique test bench for the next generation CUORE experiment, which is currently in the construction phase. The excellent performance obtained with CUORICINO demon- strates the feasibility of a ton scale bolometric experiment aiming at the investigation of mββ in the inverted mass hierarchy range.
In this chapter the experimental set-up, the analysis procedures, the detector per- formance and the physics results obtained by CUORICINO will be presented.